Jurkat cells are accepted model cells for primary human T lymphocytes, for example, in medical research. Their growth to tissue-like cell densities (up to 100 x 10(6) cells/mL(capsule)) in semi-permeable (molecular weight cut off <10,000 Da) sodium cellulose sulfate/poly(diallyldimethylammonium chloride) polyelectrolyte capsules has previously been shown by us [Werner et al. (2013). Use of the mitochondria toxicity assay for quantifying the viable cell density of microencapsulated jurkat cells. Biotechnol Prog 29(4): 986-993]. Herein, we demonstrate that encapsulation can be used to retain the cells in continuously operated bioreactors, which opens new possibilities for research, for example, the use of Jurkat cells in pulse response experiments under steady state conditions. Two reactor concepts are presented, a fluidized and a fixed bed reactor. A direct comparison of the growth kinetics in batch and repeated batch spinner cultivations, that is, under conditions where both encapsulated and non-encapsulated cells can be cultivated under otherwise identical conditions, showed that maximum specific growth rates were higher for the encapsulated than for the non-encapsulated cells. In the subsequent batch and repeated batch bioreactor experiments (only encapsulated cells), growth rates were similar, with the exception of the fixed bed batch reactor, where growth kinetics were significantly slower. Concomitantly, a significant fraction of the cells towards the bottom of the bed were no longer metabolically active, though apparently not dead. In the repeated batch fluidized bed reactor cellular division could be maintained for more than two weeks, albeit with a specific growth rate below the maximum one, leading to final cell densities of approximately 180 x 10(6) cell/g(capsule). At the same time, the cell cycle distribution of the cells was shifted to the S and G2/M phases. (C) 2014 Wiley Periodicals, Inc.